06:08 <Peng_Fan> marex: Jaehoon is doing the MMC work now.

11:03 <sjg1> marex: I pushed a new tree. It handles the -n parameter now

11:03 <sjg1> https://www.irccloud.com/pastebin/ND4a7xZK/

12:03 <marex> sjg1: I probably wont be able to look at it today

13:46 <rfs613> I have to say, I was kind of surprised to see -n used to specify the *.cfg file.

13:47 <rfs613> from the help text output, I would not have expected this.

13:47 <rfs613> but it does seem to be quite common, now that I know about it ;-)

13:48 <rfs613> the description of -f seems a litle bit more obvious

14:23 <sjg1> rfs613: Yes, same here. What is strange about imx8 is that it uses -n to specify a .cfg file which itself contains the filename of the input data. From what I understand from Marek (and from scanning the mkimage source code) it actually ignores the input data

14:23 <sjg1> rfs613: Perhaps when it is eventually converted to binman it will be possible to tidy things up a bit

14:28 <sjg1> marex: OK, will wait to hear

14:59 <rfs613> sjg1: hmm, so the ones I looked at (mtk_image and kwbimage) seem to use -d to specify the input data, and -n for the config file.

23:35 <sjg1> rfs613: That makes more sense to me. It is possible that we (Marek & I) misunderstand imx8, but we'll see

23:56 <marex> sjg1: I never worked with imx8

23:56 <marex> sjg1: I think you have imx8m in mind, it's a different soc

23:57 <marex> but the imximage format goes back like 15 years or so, it evolved, but not by much

23:58 <marex> the config file itself goes back to DCD specification which was already present in imx5 btw

23:58 <marex> so that's a legacy supreme